Deep learning-based asymmetric text hashing method

A deep learning and asymmetric technology, applied in unstructured text data retrieval, text database clustering/classification, other database retrieval, etc., can solve problems such as different expression methods and inability to accurately retrieve effective information, and achieve improved efficiency effect

Inactive Publication Date: 2017-05-31
SUN YAT SEN UNIV
View PDF7 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] However, the current text hashing learning method only uses a single hash function to model both the retrieval and the retrieved text. However, in real applications, t

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Deep learning-based asymmetric text hashing method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0032] like figure 1 As shown, an asymmetric text hashing method based on deep learning includes the following steps

[0033] A. Preprocessing: extract the text semantic labels of the training set, and calculate the semantic similarity between samples;

[0034] B. Calculate the expected hash code: According to the semantic label of the training set sample, calculate the binary hash code of the training set sample, the binary code guarantees to maintain the performance of the best semantic preservation on the training set;

[0035] C. Calculate the hash code of the text: input the text into the corresponding neural network, and calculate the hash code corresponding to the text;

[0036] D. Optimize neural network parameters: Calculate the deviation between the hash code output by the neural network and the expected hash code, and train the neural network parameters through the backpropagation algorithm.

[0037] The above steps are specifically:

[0038] Step A: Extract the ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a deep learning-based asymmetric text hashing method. According to the method, the semantic consistency of hash codes is maintained by using a difference value between binary hash code similarity and semantic similarity among minimized texts, so that the binary codes have similar information retention; a heterogeneous neural network is used for carrying out hash learning respectively on retrieving texts and retrieved texts, so that the efficiency of text hash learning is improved.

Description

technical field [0001] The present invention relates to the fields of text information retrieval and deep learning, and more specifically, relates to an asymmetric text hashing method based on deep learning. Background technique [0002] With the development of the Internet, various websites emerge in endlessly. As a general text retrieval tool, search engine has become an important entrance linking users and websites. In traditional search engines, users need to input some keywords, and the search engine uses indexing technology to retrieve articles that are highly relevant to user requests based on these keywords. In recent years, with the advancement of natural language processing technology and the development of social network and multimedia technology, there has been a need for richer text retrieval, such as: question answering system, voice assistant, recommendation system, etc. These requirements make the text retrieval method based on keyword matching start to enc...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06F17/27G06F17/30
CPCG06F16/35G06F16/951G06F40/30
Inventor 陈正梁潘嵘
Owner SUN YAT SEN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap